Abstract
Background
Warfarin remains a widely used anticoagulant for long-term thromboprophylaxis, but its narrow therapeutic index and substantial interindividual variability increase the risk of bleeding complications. Functional variants in vascular endothelial growth factor A (VEGFA) may therefore influence vascular integrity during anticoagulant therapy. This study investigated the association between functional VEGFA polymorphisms and warfarin-associated bleeding in a Saudi cohort.
Methods
In this retrospective cohort study, 161 Saudi patients receiving stable maintenance warfarin therapy were genotyped for five VEGFA single-nucleotide polymorphisms (SNPs): rs699947, rs833069, rs2010963, rs35410204, and rs866236. Clinical characteristics and bleeding events during warfarin therapy were obtained from electronic medical records. Bleeding status was determined from clinically documented bleeding events during warfarin therapy that required medical evaluation, hospitalization, or blood transfusion. Genotype-phenotype associations were assessed using chi-square analysis and a multivariable logistic regression model.
Results
Bleeding occurred in 39% of the patients (63/161). Among the five VEGFA variants studied, rs699947 and rs833069 showed associations with bleeding in bivariate analysis. The rs699947 A/A genotype and rs833069 T/T genotype were less frequent in patients with bleeding than in those without bleeding (6.7% vs 25.5%, p = 0.009; and 16.4% vs 37.5%, p = 0.013, respectively). The rs833069 association did not survive Bonferroni correction for multiple testing and is therefore considered exploratory. Following adjustment for age, sex, body mass index, international normalized ratio, systolic blood pressure, and diastolic blood pressure, rs699947 remained significantly associated with bleeding. For rs833069, the overall genotype association remained statistically significant; however, individual genotype comparisons were not statistically significant. For rs699947, the A/A genotype was associated with lower odds of bleeding compared with the C/C genotype (OR 0.249, 95% CI 0.062–0.999; p = 0.050). No significant associations were observed for rs2010963, rs35410204, or rs866236.
Conclusion
In this exploratory retrospective cohort study, VEGFA rs699947 and rs833069 showed possible associations with bleeding during warfarin therapy in Saudi patients. These findings provide preliminary evidence that VEGFA polymorphisms may contribute to bleeding susceptibility during warfarin therapy. Given the modest sample size, these findings should be interpreted cautiously and require validation in larger prospective studies.
Keywords: VEGFA, warfarin, bleeding risk, pharmacogenetics, Saudi population, rs699947, rs833069
Plain Language Summary
Warfarin is a commonly used blood thinner that helps prevent harmful blood clots. However, some people who take warfarin develop bleeding complications, even when their treatment is carefully monitored. Most genetic studies of warfarin have focused on genes that affect warfarin dose, but other genes may also influence bleeding risk.
In this study, the researchers looked at five common changes in the VEGFA gene in Saudi patients receiving stable warfarin treatment. The VEGFA gene helps regulate blood vessel health, repair, and leakage. The team compared patients who had bleeding during warfarin therapy with those who did not.
Bleeding occurred in 63 of the 161 patients. Two VEGFA gene changes, called rs699947 and rs833069, showed possible links with bleeding. One genetic pattern in rs699947 appeared less common in patients who experienced bleeding, suggesting it may be linked with lower bleeding risk. The other three VEGFA gene changes studied did not show clear associations with bleeding.
These findings suggest that genes involved in blood vessel function may help explain why some patients are more likely to bleed during warfarin treatment. However, this was an exploratory study with a modest sample size, so the results need to be confirmed in larger studies before they can be used in clinical practice.
Introduction
Warfarin, a coumarin-derived oral anticoagulant, remains a cornerstone of long-term thromboprophylaxis and continues to be widely prescribed worldwide, with an estimated 25 to 30 million patients receiving warfarin therapy annually.1 Despite the growing use of direct oral anticoagulants (DOACs), warfarin remains the preferred option in selected patient groups, particularly those with mechanical heart valves, severe renal impairment, or antiphospholipid syndrome, in whom DOACs may be contraindicated or less suitable.2 Its continued global use is further supported by its cost-effectiveness, broad availability, and the ability to monitor therapeutic response using the international normalized ratio (INR).3 Bleeding is among the most serious adverse effects of warfarin and may result in hospital admission, blood transfusion, and death, with an estimated annual mortality rate of 0.1–0.3%.4,5
The VEGFA (vascular endothelial growth factor A) gene is a member of the PDGF/VEGF growth factor family that promotes endothelial migration and vascular permeability.6 The VEGFA gene is highly polymorphic, with at least 30 reported single-nucleotide polymorphisms (SNPs).7 Several of these variants have been associated with altered gene expression, mRNA stability, and circulating VEGFA levels.8,9 Such functional variation may influence endothelial integrity and vascular repair capacity, thereby modifying susceptibility to bleeding even when warfarin therapy is clinically monitored and optimized.9
Most pharmacogenomic studies of warfarin have focused on canonical variants in CYP2C9 and VKORC1, which are established determinants of dose requirements and, in some settings, bleeding-related outcomes.10–12 In our previous study of Saudi patients receiving warfarin, VKORC1 and CYP2C9 polymorphisms were significantly associated with warfarin dose adjustment, but not with warfarin-associated bleeding.10 Together, these findings support the importance of pharmacogenetic variability in this population, while also suggesting that additional pathways beyond canonical dose-related genes may contribute to bleeding susceptibility. Because interindividual variability in bleeding risk is not fully explained by CYP2C9 and VKORC1, pathways related to vascular integrity and endothelial repair warrant investigation.13 The Greater Middle East Variome demonstrated substantial genetic diversity across regional populations, and the Qatar genome project established a population-specific genomic resource for precision medicine in the Middle East.14,15 In this context, VEGFA variants such as rs699947, rs2010963, and rs833069 are plausible candidates for population-level association studies in Arab cohorts. In addition to these commonly investigated variants, rs35410204 and rs866236 have also been evaluated in studies examining VEGFA-related vascular phenotypes and bleeding susceptibility. Although evidence regarding these variants remains limited and inconsistent across different populations, they were included in the present study to provide a more comprehensive evaluation of VEGFA genetic variation in relation to warfarin-associated bleeding.
To our knowledge, no previous study has examined the association between VEGFA variants and warfarin-associated bleeding in Arab or Middle Eastern populations. Pharmacogenomic research from the region has focused predominantly on implementation, education, and established pharmacogenes, rather than on novel bleeding-related candidates such as VEGFA.16 This represents an important knowledge gap, particularly because regional warfarin pharmacogenomic studies have largely focused on dose prediction using canonical genes such as CYP2C9 and VKORC1, with limited attention to genetic determinants of bleeding susceptibility. Because VEGFA maintains endothelial integrity, regulates vascular permeability, and promotes vascular repair, genetic variation within VEGFA represents a biologically plausible determinant of bleeding risk during anticoagulant therapy. Building on prior evidence linking VEGFA variants to vascular phenotypes and to bleeding during anticoagulation,17,18 we explored whether functional VEGFA variants are associated with bleeding during warfarin therapy in a Saudi cohort. Specifically, this study examined the association between selected VEGFA polymorphisms rs699947, rs2010963, rs833069, rs35410204, and rs866236 and bleeding during warfarin therapy, extending our previous analysis of the same Saudi warfarin cohort,10 in which VKORC1 and CYP2C9 variants were associated with dose adjustment.
Materials and Methods
Ethics Statement and Patient Recruitment
This retrospective cohort study was conducted in Saudi adult patients receiving maintenance warfarin therapy. Bleeding status was determined from clinically documented bleeding events during warfarin therapy that required medical evaluation, hospitalization, or blood transfusion. Patients were recruited between June 2016 and August 2017. The study was conducted at King Faisal Specialist Hospital and Research Centre (KFSH&RC), Riyadh, Saudi Arabia, following approval from the Office of Research Affairs at KFSH&RC (ORA: 2161251). All participants provided written informed consent for study participation and for the use of blood samples for genetic analysis. The study was conducted in accordance with the Declaration of Helsinki (2013 revision) and relevant institutional guidelines for human genomic research.
A total of 161 patients who met the inclusion criteria were included in the analysis. Eligible participants were adult Saudi patients attending the Heart Centre who had been receiving a stable warfarin maintenance dose for at least 1 month. Patients with bleeding secondary to major trauma, surgical intervention, or invasive medical procedures were excluded to minimize confounding.
Clinical and Demographic Data Collection
Clinical and demographic data, indication for warfarin therapy, concomitant medications, and documented bleeding events occurring during warfarin treatment were obtained retrospectively from electronic medical records. Patients were categorized into two groups based on the occurrence of bleeding events during warfarin therapy. The bleeding group included patients who experienced a clinically documented bleeding episode recorded in their medical records during the treatment period. Bleeding events were identified through review of hospital records and physician documentation, including events that required medical evaluation, hospitalization, or blood transfusion. The non-bleeding group consisted of patients who received warfarin therapy but did not experience any documented bleeding complications.
Genotyping of VEGFA Variants
Genomic DNA was extracted from whole-blood samples using the QIAamp DNA Mini Kit (Qiagen, Hilden, Germany). DNA quantity and quality were assessed using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, Malaysia). Five candidate VEGFA single-nucleotide polymorphisms (SNPs) were selected for analysis: rs699947 (−2578C>A), rs833069, rs2010963 (+405G>C), rs35410204, and rs866236. Variant annotation and nomenclature were based on the NCBI VEGFA reference transcript NM_003376.5 and RefSeqGene NG_008732.1. Genotyping was performed at LGC Genomics Laboratory (Berlin, Germany) using the Kompetitive Allele-Specific PCR (KASP™) genotyping system (LGC Biosearch Technologies, UK), as previously described.19 SNP genotyping showed a high success rate of 96–100% (96% for rs699947 and rs866236, 97% for rs35410204, 98% for rs833069, and 100% for rs2010963).
Statistical Analysis
Hardy-Weinberg equilibrium (HWE) was assessed for each SNP using the chi-square test. Genotype and allele frequencies were calculated for all VEGFA variants. Associations between VEGFA polymorphisms and bleeding were evaluated using the chi-square test or Fisher’s exact test, as appropriate.
To account for multiple testing across the five VEGFA SNPs examined in the primary analysis, a Bonferroni-corrected significance threshold of p ≤ 0.01 was applied. Multivariable logistic regression analysis was then performed for the two SNPs that showed at least a nominal association with bleeding in the univariate analysis (rs699947, which met the Bonferroni-corrected threshold, and rs833069, which was nominally significant but did not survive Bonferroni correction and was therefore retained on an exploratory basis) in order to assess their independent association with bleeding after adjustment for potential confounders, including age, sex, body mass index (BMI), international normalized ratio (INR), systolic blood pressure, and diastolic blood pressure. These variables were selected because they represent routinely collected clinical characteristics that have been reported to influence bleeding risk during warfarin therapy and were therefore considered potential confounders in the multivariable analysis. For the logistic regression analysis involving these two VEGFA SNPs, a Bonferroni-corrected significance threshold of p ≤ 0.025 was applied. Statistical analyses were performed using SPSS software version 29.0 (IBM Corp., Armonk, NY, USA).
Results
A total of 161 patients were included in the analysis, of whom 63 (39.1%) experienced at least one bleeding episode during warfarin therapy. Demographic and clinical characteristics of patients with and without bleeding are summarized in Table 1. No statistically significant differences were observed between the two groups for age, body mass index (BMI), pulse rate, systolic blood pressure, diastolic blood pressure, daily warfarin dose, or international normalized ratio (INR) (all p > 0.05).
Table 1.
Demographic and Clinical Characteristics of Participants According to Bleeding Status
| Variable | Total | No Bleeding | Bleeding | p-value |
|---|---|---|---|---|
| Age (years), mean ± SD | 45.76 ± 15.01 (n = 161) | 46.11 ± 16.26 | 45.22 ± 12.92 | 0.715 |
| Female, n (%) | 91 (56.5) | 55 (56.1) | 36 (57.1) | 0.899 |
| Male, n (%) | 70 (43.5) | 43 (43.9) | 27 (42.9) | |
| BMI (kg/m2), mean ± SD | 30.01 ± 10.49 (n = 152) | 29.13 ± 7.03 | 30.12 ± 7.64 | 0.148 |
| Pulse (beats/min), mean ± SD | 78.42 ± 13.39 (n = 126) | 78.92 ± 12.44 | 77.60 ± 14.91 | 0.593 |
| Systolic blood pressure (mmHg), mean ± SD | 120.99 ± 16.00 (n = 129) | 121.34 ± 16.02 | 120.44 ± 16.10 | 0.756 |
| Diastolic blood pressure (mmHg), mean ± SD | 67.80 ± 10.17 (n = 129) | 68.14 ± 10.77 | 67.26 ± 9.24 | 0.634 |
| Warfarin dose (mg/day), mean ± SD | 5.94 ± 2.79 (n = 161) | 5.98 ± 3.03 | 5.90 ± 2.39 | 0.857 |
| INR, mean ± SD | 2.51 ± 0.70 (n = 161) | 2.51 ± 0.73 | 2.50 ± 0.67 | 0.912 |
Notes: Data are presented as mean ± SD or n (%). The total sample size for each variable reflects the number of participants with available data. Bleeding was defined as a bleeding event resulting in medical evaluation, hospitalization, or blood transfusion.
Abbreviations: BMI, body mass index; INR, international normalized ratio.
Associations between categorical clinical variables and bleeding status are presented in Table 2. No significant associations were observed between bleeding and sex, warfarin indication, comorbidities, history of falls, warfarin non-compliance, or concomitant medication use (all p > 0.05). Borderline associations were observed for rheumatic heart disease (p = 0.057) and a history of falls (p = 0.093), but neither reached statistical significance.
Table 2.
Associations of Categorical Clinical Variables with Bleeding Status
| Factor | Chi-Square value | p value |
|---|---|---|
| Sex (female vs male) | 0.016 | 0.899 |
| Indication: atrial fibrillation | 0 | 1 |
| Indication: pulmonary embolism | 0.415 | 0.519 |
| Indication: deep vein thrombosis | 1.399 | 0.237 |
| Indication: mechanical heart valve | 1.876 | 0.171 |
| Indication: myocardial infarction (post-MI) | 0.002 | 0.968 |
| Indication: stroke (pre-warfarin) | 0.697 | 0.404 |
| Indication: systemic lupus erythematosus (SLE) | 0.043 | 0.835 |
| Indication: antiphospholipid syndrome (APS) | 2.506 | 0.113 |
| Indication: congestive heart failure (CHF) | 1.565 | 0.211 |
| Indication: post-Fontan procedure | 0.973 | 0.324 |
| Indication: rheumatic heart disease (RHD) | 3.619 | 0.057 |
| Indication: protein C or S deficiency | 0.647 | 0.421 |
| Indication: other | 0.1 | 0.751 |
| Warfarin non-compliance | 1.302 | 0.254 |
| History of falls | 2.819 | 0.093 |
| Diabetes (comorbidity) | 0.01 | 0.922 |
| Hypertension (comorbidity) | 0.084 | 0.772 |
| Renal impairment (chronic kidney disease) | 2.186 | 0.139 |
| Use of hypoglycemic medications (insulin or oral agents) | 0.032 | 0.857 |
| Statin use | 0.569 | 0.451 |
| Use of other anticoagulants (concomitant) | 0 | 1 |
| NSAID use | 0.067 | 0.796 |
| Thyroid hormone therapy | 0 | 1 |
| ACE inhibitor or ARB use | 0 | 1 |
| Antibiotic use (any, aside from metronidazole) | 1.445 | 0.229 |
| Calcium channel blocker use | 0.044 | 0.833 |
| Beta blocker use | 2.294 | 0.13 |
| Proton pump inhibitor (PPI) use | 0.07 | 0.792 |
| Other concomitant medications | 1.111 | 0.292 |
| Use of any warfarin-interacting drugs | 0.017 | 0.895 |
Note: Associations were assessed using the chi-square test.
Abbreviations: ACE, angiotensin-converting enzyme; ARB, angiotensin receptor blocker; NSAID, nonsteroidal anti-inflammatory drug.
All of the investigated SNPs were consistent with Hardy–Weinberg equilibrium in the study population (all p > 0.05; Supplementary Table 1). The associations between VEGFA polymorphisms and bleeding during warfarin therapy are shown in Table 3. A significant difference in genotype distribution was identified for rs699947, which met the Bonferroni-corrected threshold (p ≤ 0.01), whereas rs833069 showed a nominally significant difference that did not survive Bonferroni correction. For rs699947, the A/A genotype was less frequent among patients with bleeding than among those without bleeding (6.7% vs 25.5%; p = 0.009), whereas the A/C genotype was more frequent in the bleeding group (58.3% vs 41.5%). Similarly, for rs833069, the T/T genotype was less frequent in patients with bleeding than in those without bleeding (16.4% vs 37.5%; p = 0.013), while the T/C genotype was more common in the bleeding group (59.0% vs 39.6%). No significant differences in genotype distribution were observed for rs2010963, rs35410204, or rs866236 (all p > 0.05).
Table 3.
Association of VEGFA Polymorphisms with Bleeding During Warfarin Therapy
| SNP | Genotype | Total (n, %) | Bleeding (n, %) | No Bleeding (n, %) | p value |
|---|---|---|---|---|---|
| rs699947 | A/A | 28 (18.2%) | 4 (6.7%) | 24 (25.5%) | 0.009 |
| A/C | 74 (48.1%) | 35 (58.3%) | 39 (41.5%) | ||
| C/C (ref) | 52 (33.8%) | 21 (35.0%) | 31 (33.0%) | ||
| rs833069 | T/T | 46 (29.3%) | 10 (16.4%) | 36 (37.5%) | 0.013 |
| T/C | 74 (47.1%) | 36 (59.0%) | 38 (39.6%) | ||
| C/C (ref) | 37 (23.6%) | 15 (24.6%) | 22 (22.9%) | ||
| rs2010963 | G/G (ref) | 55 (34.6%) | 16 (25.8%) | 39 (40.2%) | 0.145 |
| C/G | 76 (47.8%) | 35 (56.5%) | 41 (42.3%) | ||
| C/C | 28 (17.6%) | 11 (17.7%) | 17 (17.5%) | ||
| rs35410204 | T/T (ref) | 123 (79.4%) | 47 (78.3%) | 76 (80.0%) | 0.934 |
| C/T | 29 (18.7%) | 12 (20.0%) | 17 (17.9%) | ||
| C/C | 3 (1.9%) | 1 (1.7%) | 2 (2.1%) | ||
| rs866236 | T/T (ref) | 23 (15.0%) | 10 (16.9%) | 13 (13.8%) | 0.509 |
| C/T | 66 (43.1%) | 22 (37.3%) | 44 (46.8%) | ||
| C/C | 64 (41.8%) | 27 (45.8%) | 37 (39.4%) |
Note: Genotype distributions were compared between patients with bleeding and those without bleeding using the chi-square test (df = 2). Percentages in the bleeding and No bleeding columns are calculated within each bleeding-status group. Percentages in the Total column are calculated using the overall number of genotyped participants for each SNP.
Abbreviations: SNP, single nucleotide polymorphism; VEGFA, vascular endothelial growth factor A; df, degrees of freedom.
Multivariable logistic regression analysis was performed for the two VEGFA variants that showed at least a nominal association in the bivariate analysis (Table 4). Following adjustment for age, sex, body mass index, international normalized ratio, systolic blood pressure, and diastolic blood pressure, rs699947 remained significantly associated with bleeding (overall SNP effect p = 0.025). For rs833069, the overall genotype association remained statistically significant (p = 0.021); however, the individual genotype comparisons were not statistically significant. For rs699947, the A/A genotype was associated with lower odds of bleeding compared with the C/C genotype (OR 0.249, 95% CI 0.062–0.999; p = 0.050), whereas the C/A genotype was not significantly associated with bleeding (OR 1.609, 95% CI 0.699–3.700; p = 0.264). For rs833069, although the overall SNP effect was significant, neither the T/T genotype (OR 0.369, 95% CI 0.116–1.177; p = 0.091) nor the T/C genotype (OR 1.662, 95% CI 0.665–4.153; p = 0.278) showed a statistically significant association when compared individually with the C/C genotype.
Table 4.
Multivariable Logistic Regression Analysis of VEGFA Polymorphisms Associated with Bleeding During Warfarin Therapy
| Gene | SNP (rsID) | Comparison | OR (95% CI) | p-value |
|---|---|---|---|---|
| VEGFA | rs699947 | C>A | — | 0.025 |
| C/C (reference) | 1 | — | ||
| C/A vs C/C | 1.609 (0.699–3.700) | 0.264 | ||
| A/A vs C/C | 0.249 (0.062–0.999) | 0.050 | ||
| rs833069 | C>T | — | 0.021 | |
| C/C (reference) | 1 | — | ||
| T/T vs C/C | 0.369 (0.116–1.177) | 0.091 | ||
| T/C vs C/C | 1.662 (0.665–4.153) | 0.278 |
Notes: Logistic regression model adjusted for age, sex, body mass index (BMI), international normalized ratio (INR), systolic blood pressure, and diastolic blood pressure. Bold values indicate statistically significant associations according to the Bonferroni-corrected significance threshold.
Abbreviations: VEGFA, vascular endothelial growth factor A; SNP, single nucleotide polymorphism; OR, odds ratio; CI, confidence interval.
Discussion
In the current exploratory retrospective cohort study, 161 Saudi patients receiving maintenance warfarin therapy were analyzed, of whom 39.1% experienced bleeding events. No significant associations were observed between bleeding and age, body mass index, blood pressure, INR, warfarin dose, clinical indications, comorbidities, or concomitant medications. Two VEGFA variants, rs699947 and rs833069, showed associations with bleeding in the present dataset. The rs699947 A/A genotype was less frequent among patients with bleeding and remained associated with lower odds of bleeding after adjustment. For rs833069, genotype distributions differed at a nominal significance level (p = 0.013) that did not survive Bonferroni correction, and the overall SNP effect remained significant in the multivariable model, although individual genotype comparisons were not statistically significant after adjustment. These findings should be interpreted cautiously and considered exploratory.
When interpreted in the context of our previous work,10 in which VKORC1 and CYP2C9 polymorphisms were significantly associated with warfarin dose adjustment, the two studies suggest that pharmacogenetic variability in this population may extend beyond canonical dose-related genes. However, whereas our prior work addressed warfarin maintenance dose, the present study explores bleeding as a broader clinical outcome and should therefore be considered hypothesis-generating.
Previous studies have established that polymorphisms in the CYP2C9 gene, which plays a major role in the metabolic clearance of the S-enantiomer of warfarin, and in the VKORC1 gene, which influences sensitivity to the vitamin K epoxide reductase complex, are important determinants of warfarin dose requirements and treatment response.10–12 Variants in CYP2C9 and VKORC1, together with age, BMI, diet, and comorbidities, explain the interindividual variability in warfarin dose requirements.20 Incorporation of genetic information into warfarin dosing algorithms has been shown to improve dose prediction and reduce the incidence of warfarin-related adverse outcomes.1 In addition, recent meta-analyses and randomized controlled trials have shown that CYP2C9 and VKORC1 genotyping prior to therapy can facilitate dose optimization, reduce drug-related hospitalization, and improve cost-effectiveness.21,22
However, a substantial proportion of warfarin-associated bleeding remains unexplained, indicating that additional genetic factors may influence bleeding susceptibility independently of the classical pharmacokinetic and pharmacodynamic pathways.3 In this context, the VEGFA gene is of particular interest because of its central role in angiogenesis, endothelial repair, vascular permeability, and maintenance of vascular homeostasis.7 Variation in VEGFA may therefore affect vascular resilience during anticoagulation and modify the risk of bleeding even when warfarin therapy is clinically monitored.
The present study showed that the rs699947 variant was significantly associated with bleeding risk during warfarin therapy, with the C allele showing an apparent association with greater bleeding susceptibility and the A allele showing an apparent protective association in this cohort.
This finding is consistent with the possibility that genetic variation in angiogenesis-related pathways may influence bleeding risk independently of warfarin dose and INR. Previous studies have linked rs699947 to altered VEGFA expression and vascular responses.23,24 Such effects may influence endothelial stability and vascular permeability, thereby modifying bleeding susceptibility during anticoagulant therapy. Beyond the anticoagulation setting, the rs699947 C allele has been associated with idiopathic heavy menstrual bleeding,25 while cardiovascular studies have linked the same allele to maladaptive vascular remodeling and endothelial dysfunction.23,24 Conversely, the A allele has been associated with reduced VEGF expression and more stable vascular phenotypes.23,25 Taken together, these findings suggest that rs699947 may warrant further investigation as a variant potentially related to VEGFA-mediated vascular regulation.
Although bleeding was not examined as an outcome, a recent Saudi cohort study identified rs699947 as a functionally relevant angiogenesis-related VEGFA variant with context-dependent vascular effects, supporting the biological plausibility of allele-dependent regulation of VEGF signaling in vascular integrity.26 This lends additional support to the hypothesis that VEGFA-related endothelial mechanisms may contribute to bleeding susceptibility during warfarin therapy.
The rs833069 variant showed a nominal association with bleeding in the present cohort; although the overall SNP effect reached significance in the multivariable model, the univariate association did not survive Bonferroni correction. Although individual genotype comparisons did not reach statistical significance in the multivariable model, the bivariate analysis showed a different genotype distribution between patients with and without bleeding, suggesting a possible role for this variant in bleeding susceptibility. This represents a potentially important finding, as rs833069 has been less extensively studied than promoter-region VEGFA variants such as rs699947 and rs2010963. Nevertheless, this intronic SNP has been implicated in angiogenic signaling and vascular function.9,27 rs833069 lies within a haplotype block that includes the promoter variant rs833061, and these linked variants may jointly influence VEGFA transcription, providing a plausible functional basis for the observed association.28 In addition, VEGFA polymorphisms have been associated with intracerebral hemorrhage risk in patients with brain arteriovenous malformations, supporting a broader role for VEGFA-mediated vascular instability in bleeding-related phenotypes.27 The rs833069 variant has also been linked to regulation of the VEGF pathway and vascular remodeling in age-related macular degeneration.29,30 Although direct evidence linking rs833069 to warfarin-associated bleeding has been lacking, the present findings suggest that this locus may warrant further investigation in relation to vascular fragility and endothelial repair.
Previous work in a Korean cohort identified significant associations of rs35410204 and rs866236 with warfarin-associated bleeding in patients with mechanical heart valves, further supporting a potential role for VEGFA-mediated vascular mechanisms in anticoagulant safety.18 In that study, the rs35410204 C/C genotype was associated with an approximately 10-fold increase in bleeding risk, whereas wild-type homozygosity at rs866236 was associated with an approximately 2.9-fold increase in risk after adjustment for clinical covariates.18 In contrast, neither rs35410204 nor rs866236 was significantly associated with bleeding in our Saudi cohort. In our dataset, the rs35410204 C allele was uncommon, and the C/C genotype was rare, which may have reduced power to detect an effect; by contrast, rs866236 showed a markedly different genotype distribution, with the C allele being common in this cohort. These between-study differences are consistent with population-specific genetic architecture and suggest that the VEGFA markers associated with bleeding risk may differ across ethnic groups. A further explanation is that linkage disequilibrium patterns across the VEGFA locus may differ between East Asian and Middle Eastern populations, such that rs35410204 and rs866236 may tag functional regulatory variation in Koreans, whereas rs699947 and rs833069 may better capture the relevant signal in our cohort.31,32
Study Limitations
This study has several limitations. First, the retrospective design limited control over phenotype characterization and covariate completeness. In addition, several established clinical predictors of bleeding, including renal function, liver function, serum albumin concentration, previous bleeding history, and other potentially relevant clinical variables, were not consistently available because of the retrospective study design and therefore could not be included in the multivariable analysis. Consequently, residual confounding cannot be excluded. Second, bleeding was defined using clinically documented events extracted from medical records rather than a prospectively adjudicated standardized bleeding classification, which may have introduced heterogeneity and outcome misclassification. Third, no functional validation was performed, such as assessment of VEGFA gene expression or circulating VEGFA levels; therefore, the biological consequences of the identified variants could not be directly confirmed. Fourth, linkage disequilibrium patterns across the VEGFA locus were not comprehensively characterized. As a result, rs699947 and rs833069 may not themselves be causal variants but may instead be in linkage disequilibrium with other functional variants within a larger haplotype block. Fifth, the relatively modest sample size may have limited statistical power and restricted the number of covariates that could be included in the multivariable analysis. Accordingly, the findings should be interpreted as exploratory and require validation in larger independent cohorts. In addition, the canonical warfarin pharmacogenes CYP2C9 and VKORC1 were not entered into the present multivariable model. These variants were genotyped and analysed in the same Saudi cohort in our previous study, in which they were significantly associated with warfarin maintenance dose but showed no significant association with warfarin-associated bleeding.10 Because they were not associated with the bleeding phenotype in this population, they were not treated as confounders of the present outcome, which supports their exclusion from the model. Nevertheless, future studies jointly modelling VEGFA polymorphisms and established warfarin pharmacogenes in larger cohorts may provide a more comprehensive assessment of bleeding risk during warfarin therapy.
Conclusions
In this exploratory retrospective cohort study, VEGFA rs699947 and rs833069 showed possible associations with bleeding during warfarin therapy in Saudi patients. These findings extend our previous pharmacogenetic work in the same population beyond canonical dose-related genes, but should be interpreted cautiously given the retrospective design, modest sample size, broad bleeding phenotype, and lack of independent replication. Further studies in larger, prospectively characterized cohorts are needed to determine whether VEGFA variation independently contributes to bleeding susceptibility during warfarin therapy.
Funding Statement
This study was supported by the College of Medicine, Office of Research, Alfaisal University.
Use of Generative AI
The authors used ChatGPT (OpenAI, GPT-4o) to assist with language editing, grammar improvement, and refinement of manuscript presentation. All scientific content, data interpretation, conclusions, and final editorial decisions were reviewed and approved by the authors. The authors take full responsibility for the content of this manuscript.
Abbreviations
BMI, Body mass index; CI, Confidence interval; CYP2C9, Cytochrome P450 family 2 subfamily C member 9; DOACs, Direct oral anticoagulants; HWE, Hardy–Weinberg equilibrium; INR, International normalized ratio; KFSH&RC, King Faisal Specialist Hospital and Research Centre; OR, Odds ratio; PCR, Polymerase chain reaction; SNP, Single-nucleotide polymorphism; VEGF, Vascular endothelial growth factor; VEGFA, Vascular endothelial growth factor A; VKORC1, Vitamin K epoxide reductase complex subunit 1.
Data Sharing Statement
The datasets generated and/or analyzed during the current study are not publicly available because of institutional and ethical restrictions. De-identified data may be made available from the corresponding author on reasonable request, subject to approval by the relevant institutional ethics committee and applicable data-sharing regulations.
Ethics Approval and Informed Consent
This study was approved by the Office of Research Affairs at King Faisal Specialist Hospital and Research Centre, Riyadh, Saudi Arabia (ORA No. 2161251). All participants provided written informed consent for participation and for the use of blood samples for genetic analysis. The study was conducted in accordance with the principles of the Declaration of Helsinki and relevant institutional guidelines.
Author Contributions
All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.
Disclosure
The authors report no conflicts of interest in this work.
References
- 1.Kaye JB, Schultz LE, Steiner HE, Kittles RA, Cavallari LH, Karnes JH. Warfarin pharmacogenomics in diverse populations. Pharmacotherapy. 2017;37:1150–11. doi: 10.1002/phar.1982 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Cheng A, Malkin C, Briffa NP. Antithrombotic therapy after heart valve intervention: review of mechanisms, evidence and current guidance. Heart. 2023;110:87–93. doi: 10.1136/heartjnl-2022-321387 [DOI] [PubMed] [Google Scholar]
- 3.Eby C. Warfarin pharmacogenetics: does more accurate dosing benefit patients? Semin Thromb Hemost. 2012;38:661–666. doi: 10.1055/s-0032-1326789 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Gage BF, Eby C, Johnson JA, et al. Use of pharmacogenetic and clinical factors to predict the therapeutic dose of warfarin. Clin Pharmacol Ther. 2008;84:326–331. doi: 10.1038/clpt.2008.10 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Pirmohamed M. Warfarin: almost 60 years old and still causing problems. Br J Clin Pharmacol. 2006;62:509–511. doi: 10.1111/j.1365-2125.2006.02806.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Olsson AK, Dimberg A, Kreuger J, Claesson-Welsh L. VEGF receptor signalling - in control of vascular function. Nat Rev Mol Cell Biol. 2006;7:359–371. doi: 10.1038/nrm1911 [DOI] [PubMed] [Google Scholar]
- 7.Buroker NE. VEGFA rSNPs, transcriptional factor binding sites and human disease. J Physiol Sci. 2014;64:73–76. doi: 10.1007/s12576-013-0293-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Debette S, Visvikis-Siest S, Chen MH, et al. Identification of cis- and trans-acting genetic variants explaining up to half the variation in circulating vascular endothelial growth factor levels. Circ res. 2011;109:554–563. doi: 10.1161/CIRCRESAHA.111.243790 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Watson CJ, Webb NJ, Bottomley MJ, Brenchley PE. Identification of polymorphisms within the vascular endothelial growth factor (VEGF) gene: correlation with variation in VEGF protein production. Cytokine. 2000;12:1232–1235. doi: 10.1006/cyto.2000.0692 [DOI] [PubMed] [Google Scholar]
- 10.Holail J, Mobarak R, Al-Ghamdi B, Aljada A, Fakhoury H. Association of VKORC1 and CYP2C9 single-nucleotide polymorphisms with warfarin dose adjustment in Saudi patients. Drug Metab Pers Ther. 2022;37:353–359. doi: 10.1515/dmpt-2022-0108 [DOI] [PubMed] [Google Scholar]
- 11.Wadelius M, Chen LY, Downes K, et al. Common VKORC1 and GGCX polymorphisms associated with warfarin dose. Pharmacogenomics J. 2005;5:262–270. doi: 10.1038/sj.tpj.6500313 [DOI] [PubMed] [Google Scholar]
- 12.Klein TE, Altman RB, Eriksson N, et al. Estimation of the warfarin dose with clinical and pharmacogenetic data. New Engl J Med. 2009;360:753–764. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Shamieh SE, Saleem RA, Halat DH, et al. Integrating pharmacogenomics in three Middle Eastern countries’ healthcare (Lebanon, Qatar, and Saudi Arabia): current insights, challenges, and strategic directions. PLoS One. 2025;20:e0319042. doi: 10.1371/journal.pone.0319042 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Scott EM, Halees A, Itan Y, et al. Characterization of Greater Middle Eastern genetic variation for enhanced disease gene discovery. Nat Genet. 2016;48:1071–1076. doi: 10.1038/ng.3592 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Fakhro KA, Staudt MR, Ramstetter MD, et al. The Qatar genome: a population-specific tool for precision medicine in the Middle East. Hum Genome Var. 2016;3:16016. doi: 10.1038/hgv.2016.16 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Shamieh SE, Zgheib NK. Pharmacogenetics in developing countries and low resource environments. Hum Genet. 2022;141:1159–1164. doi: 10.1007/s00439-021-02260-9 [DOI] [PubMed] [Google Scholar]
- 17.Hagstrom SA, Ying GS, Pauer GJ, et al. VEGFA and VEGFR2 gene polymorphisms and response to anti-vascular endothelial growth factor therapy: comparison of age-related macular degeneration treatments trials (CATT). JAMA Ophthalmol. 2014;132:521–527. doi: 10.1001/jamaophthalmol.2014.109 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Yee J, Kim W, Chang BC, Chung JE, Lee KE, Gwak HS. Association between VEGFA gene polymorphisms and bleeding complications in patients maintaining therapeutic international normalized ratio. Pharmacogenomics. 2019;20:659–667. doi: 10.2217/pgs-2019-0005 [DOI] [PubMed] [Google Scholar]
- 19.Shamieh SE, Costanian C, Kassir R, Visvkis-Siest S, Bissar-Tadmouri N. APOE genotypes in Lebanon: distribution and association with hypercholesterolemia and Alzheimer’s disease. Per Med. 2019;16:15–23. doi: 10.2217/pme-2018-0067 [DOI] [PubMed] [Google Scholar]
- 20.Lähteenmäki J, Vuorinen AL, Lehto M, Niemi M, Forsberg MM. Pharmacogenetics of warfarin and healthcare costs - Real-world data analysis. Pharmacoepidemiol Drug Saf. 2023;32:382–386. doi: 10.1002/pds.5585 [DOI] [PubMed] [Google Scholar]
- 21.Pirmohamed M, Burnside G, Eriksson N, et al. A randomized trial of genotype-guided dosing of warfarin. New Engl J Med. 2013;369:2294–2303. doi: 10.1056/NEJMoa1311386 [DOI] [PubMed] [Google Scholar]
- 22.Verbelen M, Weale ME, Lewis CM. Cost-effectiveness of pharmacogenetic-guided treatment: are we there yet? Pharmacogenomics J. 2017;17:395–402. doi: 10.1038/tpj.2017.21 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Osadnik T, Lekston A, Bujak K, Strzelczyk JK, Poloński L, Gąsior M. The relationship between VEGFA and TGFB1 polymorphisms and target lesion revascularization after elective percutaneous coronary intervention. Dis Markers. 2017;2017:8165219. doi: 10.1155/2017/8165219 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Skrypnik D, Mostowska A, Jagodziński PP, Bogdański P. Association of rs699947 (−2578 C/A) and rs2010963 (−634 G/C) single nucleotide polymorphisms of the VEGF Gene, VEGF-A and leptin serum level, and cardiovascular risk in patients with excess body mass: a case-control study. J Clin Med. 2020;9:469. doi: 10.3390/jcm9020469 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Kaygusuz I, Gündüz NS. Investigation of vascular endothelial growth factor (VEGF) polymorphism in patients with idiopathic heavy menstrual bleeding. Arch Gynecol Obstetr. 2022;305:109–116. doi: 10.1007/s00404-021-06225-1 [DOI] [PubMed] [Google Scholar]
- 26.Almotairi R, Mir R, Almasoudi KS, Husain E, Mtiraoui N. Evaluating the impact of rs4025935, rs71748309, rs699947, and rs4646994 genetic determinants on polycystic ovary syndrome predisposition-a case-control study. Life. 2025;15:558. doi: 10.3390/life15040558 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Gong ZP, Qiao ND, Gu YX, et al. Polymorphisms of VEGFA gene and susceptibility to hemorrhage risk of brain arteriovenous malformations in a Chinese population. Acta Pharmacol Sin. 2011;32:1071–1077. doi: 10.1038/aps.2011.76 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Wu X, Xin Z, Zhang W, et al. Polymorphisms in the VEGFA promoter are associated with susceptibility to hepatocellular carcinoma by altering promoter activity. Int J Cancer. 2013;133:1085–1093. doi: 10.1002/ijc.28109 [DOI] [PubMed] [Google Scholar]
- 29.Hong N, Shen Y, Yu CY, Wang SQ, Tong JP. Association of the polymorphism Y402H in the CFH gene with response to anti-VEGF treatment in age-related macular degeneration: a systematic review and meta-analysis. Acta ophthalmologica. 2016;94:334–345. doi: 10.1111/aos.13049 [DOI] [PubMed] [Google Scholar]
- 30.Chang FY, Huang CH, Yang CH, et al. Genetics in neovascular age-related macular degeneration susceptibility and treatment response to anti-VEGF intravitreal injection: a case series study. Clin Exp Ophthalmol. 2024;52:655–664. doi: 10.1111/ceo.14388 [DOI] [PubMed] [Google Scholar]
- 31.Auton A, Brooks LD, Durbin RM, et al. A global reference for human genetic variation. Nature. 2015;526:68–74. doi: 10.1038/nature15393 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Ramaswamy S, Jain R, Naofal ME, et al. Middle Eastern Genetic variation improves clinical annotation of the human genome. J Pers Med. 2022;12:423. doi: 10.3390/jpm12030423 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The datasets generated and/or analyzed during the current study are not publicly available because of institutional and ethical restrictions. De-identified data may be made available from the corresponding author on reasonable request, subject to approval by the relevant institutional ethics committee and applicable data-sharing regulations.